White Constituents and Congressional Voting

• Author: Eric R. Hansen
• Published date: 01 July 2022
• Date: 01 July 2022
• DOI: http://doi.org/10.1177/1532673X221087159

Article

American Politics Research

2022, Vol. 50(4) 564–582

© The Author(s) 2022

Article reuse guidelines:

sagepub.com/journals-permissions

DOI: 10.1177/1532673X221087159

journals.sagepub.com/home/apr

White Constituents and Congressional Voting

Eric R. Hansen

1



Abstract

Why do some members ofCongress vote more on the extremes of theirparty than others? I argue that lawmakersrepresenting

more homogeneously white districts have greater electoral incentive to moderate their voting records, since the two parties

compete morefor support of white voters than for thesupport of minority voters. I provideevidence using roll-call votesfrom the

U.S. House and Senate.Ifind members representing more homogeneously white districtshave more moderate voting records, a

finding that holds for Democrats and Republicans. I explore two potential mechanisms: legislator responsivenessand electoral

punishment. While legislators do not seemto adjust their voting behavior in response to short-term changes in district racial

composition, morehomogeneously white districts are foundto assess larger vote share penalties on more extreme candidates in

general elections. The findings have implications for our understanding of race, representation, and electoral accountability.

Keywords

representation, race, polarization, congress

What leads some members of Congress to vote more on the

extremes of their parties than others? The two parties have po-

larized, with lawmakers holding voting records that are more

extreme on average and more homogeneous within party caucuses

than lawmakers in prior decades (e.g., Lee, 2009;McCarty et al.,

2006). However, important variation in the extremity of individual

lawmakers’records remains to be explained. In particular, con-

stituencies can serve as a moderating influence on members’votes.

Electoral considerations at home strongly influence lawmakers’

decisions (Fenno, 1977;Kingdon, 1977;Mayhew, 1974)and

constituencies often prefer more moderate representation than their

partisan representatives provide (Bafumi & Herron, 2010;Masket

& Noel, 2012; though see Ahler & Broockman, 2018). Scholars

consistently find that voters penalize lawmakers who vote farther

to the extremes of their parties (Ansolabehere et al., 2001;Canes-

Wrone et al., 2002;Carson et al., 2010;Erikson, 1971). This

literature, grounded in spatial theories of representation (Downs,

1957), focuses narrowly on voter and lawmaker ideology.

However, social group ties also inform the actions of both

voters (Achen & Bartels, 2016;Berelson et al., 1954;

Campbell et al., 1960) and lawmakers (Bratton & Haynie,

1999;Broockman, 2013). Group attachments to party coa-

litions can allow lawmakers to engage in more ideologically

extreme representation of their districts, even if that repre-

sentation is out of step with constituency preferences (Glazer

et al., 1998). This dynamic becomes evident examining the

role of race in U.S. elections. The two parties’coalitions have

increasingly sorted along racial lines (Hajnal & Lee, 2011;

Mason, 2018). When voters are committed to one party on the

basis of racial group ties, legislators in both parties have

greater latitude to vote on the extremes in office. Voters in

uncompetitive or “captured”groups are unlikely to punish

extreme in-party incumbents by supporting out-party chal-

lengers, particularly if they lack information about the

ideological distance between them and their representatives

(Frymer, 1999;Griffin & Flavin, 2007). Likewise, out-party

incumbents have little incentive to moderate their own rec-

ords to appeal to captured voters, since moderating is unlikely

to win their votes. However, to a greater extent than other

groups, whites remain split between the two parties’coali-

tions. As a consequence, more homogeneously white con-

stituencies should be more likely to vacillate in supporting

candidates of both parties, all else equal. Legislators repre-

senting such districts should be more constrained to hold

moderate voting records. If they do not, they risk homoge-

neously white constituencies shifting their support to out-

party challengers.

I provide evidence using data summarizing individual

lawmakers’roll-call votes from the 112th and 113th House

and Senate. The evidence suggests that representatives of

more homogeneously white constituencies hold more mod-

erate voting records, even when controlling for key variables

like district ideology and electoral competitiveness. I explore

1

Loyola University Chicago, Chicago, IL, USA

Corresponding Author:

Eric R. Hansen, Department of Political Science, Loyola University Chicago,

Third Floor, Coffey Hall, 1032 W. Sheridan Rd, Chicago, IL 60660, USA.

Email: ehansen4@luc.edu

two mechanisms that might explain this association: in-

cumbent responsiveness and electoral punishment. Using a

simple redistricting design to test the incumbent respon-

siveness mechanism, I find that members of the U.S. House

who came to represent more white voters after the 2012 round

of redistricting did not subsequently moderate their roll-call

votes. A test of the electoral punishment mechanism using

DIME data (Bonica, 2014) yields mixed results. I find that in

more homogeneously white constituencies, extremists suffer

larger penalties in vote shares than in more racially diverse

districts. However, the findings do not show that extreme

candidates are more likely to lose their elections as the share

of white constituents increases. The results suggest elimi-

nating incumbent responsiveness as a potential mechanism,

but leave open the possibility that electoral punishment of

extreme candidates by white constituents contributes to

lawmaker moderation.

Though primarily descriptive, the results contribute to our

understanding of the relationship between legislative voting

and race. It is well-established that issuesofrace have helped to

divide the two parties over the last half century (Carmines &

Stimson, 1989;Miller & Schofield, 2003).While studies often

examine how district racial composition affects votes on key

issues important to a minority group (e.g., Canon, 1999;

Casellas & Leal,2013), few connect district racial composition

to overall voting records. The findings suggest that studies of

electoralresponsiveness should more thoroughlyconsider how

racial coalitions of party support shape lawmakers’behavior

beyond racial and racialized issues.

Constituent Ideology, Social Groups, and

Legislative Voting

The early 21st Century finds Democrats and Republicans in

Congress historically polarized. Lawmakers cast more party-line

votes, resulting in voting records that on average appear more

extreme and more homogeneous within parties than in prior eras

(Aldrich et al., 2014;Lee, 2009). Pressure to cast more extreme

votes comes regularly from intense in-party policy demanders

(Layman et al., 2010), ideological donors (Barber et al., 2017), and

party caucus leaders (Harden & Carsey, 2012). In reaction to these

polarizing dynamics, potential candidates’personal considerations

of the costs and benefitsofservinginoffice lead moderates to

decline to run for office while ideologues fill their seats (Hall,

2019;Thomsen, 2017). The result is a Congress with few ded-

icated moderates and many strong partisans.

Even as the parties polarize in the aggregate, individual

legislators may face electoral penalties for extreme voting

behavior. Constituencies often prefer more moderate repre-

sentation than lawmakers voting on party extremes provide

(Bafumi & Herron, 2010;Masket & Noel, 2012;thoughsee

Ahler & Broockman, 2018). As a result, lawmakers may suffer

electoral penalties for voting out of line with constituency

preferences. Incumbents who vote too often on the extremes

face a higher risk of losing reelection (Ansolabehere et al.,

2001;Canes-Wrone et al., 2002;Carson et al., 2010;Erikson,

1971) or drawing challengers (Birkhead, 2015;Hogan, 2008).

Likewise,moderate candidates tend to win largervote shares in

general electionsand win office at a higher rate than their more

extreme rivals from primary elections (Hall, 2015). For many

lawmakers,electoral pressures in their homedistricts provide a

compelling motive for moderation.

Constituencies will vary in their propensity to punish ex-

treme voting behavior from their representatives. Districts in

which averageconstituency opinion is extremelyliberal would

be less likely to punishextremely liberal representativesthan a

more moderate constituency; the same logic would apply to

conservatives. Studies of constituency representation consis-

tently show a strong association between districtideology and

the voting behavior of its representative (e.g., Clinton, 2006;

Tausanovitch & Warshaw, 2013).Ideologically heterogeneous

districts also create electoral conditions that allow more ex-

treme legislators to hold office (Fiorina, 1974;Levendusky &

Pope, 2010) and give legislators leeway to vote with party

leaders (Harden & Carsey, 2012).

In addition to constitutent ideology, the representation of

social groups factors in to lawmakers’decisions about roll-

call voting (Bailey & Brady, 1998;Bishin, 2009;Miler,

2010). By social group, I mean a set of people who share

a common identity that has some relevance to politics. Social

groups may hold similar policy preferences, but need not hold

uniform ones (Karol, 2009).

A prominent line of thinking holds that parties function as

coalitions of social groups that create potential popular

majorities, enabling those groups to be represented within

government institutions (Bawn et al., 2012). Voters who

identify with party-aligned social groups tend to vote in

majorities, though not uniformly, for their party’s candidates.

Social group attachments shape voter behavior in addition to,

but independently from, voter ideology as group identity and

consciousness come to bear in vote choice (Berelson et al.,

1954;Conover, 1988). The group’s electoral support may be

premised on a small set of issues important to the group

(Miller et al., 1981). Group voters may also weigh consid-

erations beyond issues, like group norms (White et al., 2014)

or feelings of exclusion (Kuo et al., 2017). Given the inability

of many Americans to align their issue positions in an

ideologically consistent manner (Converse, 1964) and their

propensity to arrive at issue positions after forming partisan

attachments (Layman & Carsey, 2002), vote choice for many

Americans may be more an expression of group identity or

symbolic attachments than of ideological preference (Green

et al., 2002;Achen & Bartels, 2016, though see Abramowitz

& Saunders, 2006). Increasingly, Americans are socially

sorting into parties such that their group identitiesmore re-

liably predict their party identification (Mason, 2018).

As a result, legislators’voting records should reflect in part

the social group composition of their districts. While some

support comes from voters with shared ideological prefer-

ences, support will also come from social groups aligned with

Hansen 565